Septimiu E. Salcudean

Septimiu E. Salcudean

Auto-generated profile, based on publicly available information

Septimiu E. Salcudean

Department of Electrical and Computer Engineering, University of British Columbia, Vancouver, Canada

Research Focus & Metrics

Academic Career Length

46Years (1981-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
19 July 2019 in IEEE Transactions on Medical Imaging
Verified
Journal Article
Closed access
DKDavood Karimi
Davood Karimi
SSSeptimiu E. Salcudean
Septimiu E. Salcudean
The Hausdorff Distance (HD) is widely used in evaluating medical image segmentation methods. However, existing segmentation methods do not attempt to reduce HD directly. In this paper, we present novel loss functions for training convolutional neural network (CNN)-based segmentation methods with the goal of reducing HD directly. We propose three methods to estimate HD from the segmentation probability map produced by a CNN. One method makes use of the distance transform of the segmentation boundary. Another method is based on applying morphological erosion on the difference between the true and estimated segmentation maps. The third method works by applying circular/spherical convolution kernels of different radii on the segmentation probability maps. Based on these three methods for estimating HD, we suggest three loss functions that can be used for training to reduce HD. We use these loss functions to train CNNs for segmentation of the prostate, liver, and pancreas in ultrasound, magnetic resonance, and computed tomography images and compare the results with commonly-used loss functions. Our results show that the proposed loss functions can lead to approximately 18 – 45% reduction in HD without degrading other segmentation performance criteria such as the Dice similarity coefficient. The proposed loss functions can be used for training medical image segmentation methods in order to reduce the large segmentation errors.

14 October 2003 in IEEE Transactions on Robotics and Automation
Verified
Journal Article
Closed access
SDS.P. DiMaio
S.P. DiMaio
SSS.E. Salcudean
S.E. Salcudean
A methodology for estimating the force distribution that occurs along a needle shaft during insertion is described. An experimental system for measuring planar tissue phantom deformation during needle insertions has been developed and is presented. A two-dimensional linear elastostatic material model, discretised using the finite element method, is used to derive contact force information that is not directly measurable. This approach provides a method for quantifying the needle forces and soft tissue deformations that occur during general needle trajectories in multiple dimensions. The needle force distribution is used for graphical and haptic real-time simulations of needle insertion. Since the force displacement relationship is required only along the needle, a condensation technique is shown to achieve very fast interactive simulations. Stiffness matrix changes corresponding to changes in boundary conditions and material coordinate frames are performed using low-rank matrix updates.

7 August 2002 in IEEE Transactions on Robotics and Automation
Verified
Journal Article
Closed access
KHK. Hashtrudi-Zaad
K. Hashtrudi-Zaad
SSS.E. Salcudean
S.E. Salcudean
This paper first investigates the issue of transparency in time-delayed teleoperation. It then studies the advantages of employing local force feedback for enhanced stability and performance. In addition, two classes of three-channel control architectures, that are perfectly transparent under ideal conditions are introduced. The stability robustness of the proposed architectures to delays is rigorously analyzed, leading to certain bounds on force feedforward control parameters. Experimental results are included in support of the theoretical work.

1 June 2001 in The International Journal of Robotics Research
Verified
Journal Article
Closed access
KHKeyvan Hashtrudi-Zaad
Keyvan Hashtrudi-Zaad
SSSeptimiu E. Salcudean
Septimiu E. Salcudean
A large number of bilateral teleoperation control architectures in the literature have been designed based on assumed impedance models of the master and slave manipulators. However, hydraulic or heavily geared and many other manipulators cannot be properly described by impedance models. In this paper, a common four-channel bilateral control architecture designed for the above impedance models is extended to teleoperation systems with master and slave manipulators of either the admittance or impedance type. Furthermore, control parameters that provide perfect transparency under ideal conditions are found for each type of teleoperation system. Because in practice such parameters may not lead to systems that are robust to time delays and model uncertainties, an analysis of the stability and performance robustness of this very general architecture and two-channel architectures is also presented. The analysis uses the passivity-based Llewellyn two-port network absolute stability criterion, as well as bounds on the minimum and range of values of the impedance transmitted to the operator. The results of these evaluations provide design guidelines on choosing a particular control architecture and its parameters given different master and slave manipulator structures.

1 January 1991 in IEEE Transactions on Automatic Control
Verified
Journal Article
Closed access
SSS. Salcudean
S. Salcudean

7 August 2002 in IEEE Transactions on Robotics and Automation
Verified
Journal Article
Closed access
PAP. Abolmaesumi
P. Abolmaesumi
SSS.E. Salcudean
S.E. Salcudean
WZWen-Hong Zhu
Wen-Hong Zhu
MSM.R. Sirouspour
M.R. Sirouspour
SDS.P. DiMaio
S.P. DiMaio
A robot-assisted system for medical diagnostic ultrasound has been developed by the authors. The paper presents the visual servo controller used in this system. While the ultrasound transducer is positioned by a robot, the operator, the robot controller, and an ultrasound image processor have shared control over its motion. Ultrasound image features that can be selected by the operator are recognized and tracked by a variety of techniques. Based on feature tracking, ultrasound image servoing in three axes has been incorporated in the interface and can be enabled to automatically compensate, through robot motions, unwanted motions in the plane of the ultrasound beam. The accuracy of the system is illustrated through a 3-D reconstruction of an ultrasound phantom. An Internet-based robot-assisted teleultrasound system has also been demonstrated.